R Markdown

Has the COVID-19 pandemic influenced antidepressant prescribing patterns during the winteseason (September-October) across Scottish health boards?

Winter season

scotishcensusgov

#Join the data boards

I loaded september to october data boards from 2017 - 2023 to represent the freshers season then I merged the health boards #I summarise total antidepressant prescriptions per Freshers year and plotted the graph to see the trend #Looked at pre coviid , during covid and after covid trend to see if there has been any impact or association

library(tidyverse)
library(here) # directory stucture
library(gt) # tables
library(janitor) # cleaning data
library(ggplot2) # plotting graph
library(sf) # to read in map data 
library(readxl) # to read in map data
library(plotly) # to make interactive
library(viridis)
library(sf)

loading a large amount of data in a shorter time period by downloading and using the mapdfr function (data from 2017-2023)

files <- list.files(here("data", "winter_data"), pattern = "csv")
winter_data <- files %>% 
  map_dfr(~read_csv(here("data", "winter_data", .))) %>% 
clean_names()

clean up data and filter for the sections you want

filtered_winter_data <- winter_data %>% 
filter(str_starts(bnf_item_code,"0403")) %>%  #antidepressant code is 0403
  mutate(year = as.numeric(substr(paid_date_month,1,4)), month = as.numeric(substr(paid_date_month,5,6))) %>% #separates the date into years and month so that i can group winter sections
  mutate(winter_year=case_when(month == 12 ~ year + 1, 
month %in% c(1,2) ~ year) )#makes a new column to group the winter years 

filtered_winter_data <- filtered_winter_data %>% 
  unite("healthboards",hbt2014,hbt,sep = "_")#so some of my data healthboard codes were under the name hbt_2014 AND another was hbt so i had to merge the column so all the healthboard columns fall under one
filtered_winter_data$healthboards <- gsub("[NA]","",filtered_winter_data$healthboards) 
    filtered_winter_data$healthboards <-
      gsub("_","",filtered_winter_data$healthboards)#had to remove some NA characters and '_' characters

Graph 1

winter_years_data <- filtered_winter_data %>% 
  group_by(winter_year) %>% 
  summarise(total_items=sum(number_of_paid_items,na.rm = TRUE))

plot <- ggplot(winter_years_data,aes(x=winter_year,y=total_items)) +
  geom_line(linewidth=0.7,colour = "blue") +
  geom_point(size=2)+
  scale_x_continuous(breaks=2017:2023) +
  labs(title="Antidepressant Prescriptions During Winter Season",x="Year",y="Total Antidepressant Prescriptions") +
  theme_minimal()
  
ggplotly(plot)
print(plot)

#write a code talking about the zoomed in changes and reference why you dudnt go from 0
#ROUGH
population <- readxl::read_excel(here("data","population.xlsx"), skip=10) %>% 
  clean_names() %>% 
  select(x2,all_people) %>% 
  filter(!is.na(all_people)) %>% 
  rename(h_bname = "x2",hb_population = "all_people") %>% 
filter(!str_detect(hb_population,"Cells"))

filtered2_winter_data <- filtered_winter_data %>% 
  group_by(healthboards,bnf_item_code,winter_year,gp_practice) %>% 
  summarise(total_paid = sum(paid_quantity, na.rm =TRUE))

retry reading in data

filtered3_winter_data <- filtered_winter_data %>% 
  group_by(healthboards,winter_year,paid_date_month,gp_practice) %>% 
  summarise(paid_quantity = sum(paid_quantity,na.rm=TRUE)) 

SIMD <- readxl::read_excel(here("data","SIMD.xlsx")) %>% 
  clean_names() 

combined_pop_simd <- SIMD %>% 
  full_join(filtered3_winter_data,join_by(h_bcode == healthboards)) %>% 
  left_join(population,join_by(h_bname == h_bname))

retry making map

combined_pop_simd2 <- combined_pop_simd %>% 
  group_by(winter_year,h_bname,h_bcode) %>% 
    summarise(quantity_per_head = sum(paid_quantity)/mean(hb_population)) %>% 
    filter(!is.na(h_bname)) 

### map 
NHS_healthboards <- st_read(here( "data", "Week6_NHS_HealthBoards_2019.shp")) %>% 
clean_names() %>% 
  rename(h_bname = hb_name)
## Reading layer `Week6_NHS_healthboards_2019' from data source 
##   `/Users/olufimihanfaturoti/Year 3 Medicine/data-science/B251495/data/Week6_NHS_healthboards_2019.shp' 
##   using driver `ESRI Shapefile'
## Simple feature collection with 14 features and 4 fields
## Geometry type: MULTIPOLYGON
## Dimension:     XY
## Bounding box:  xmin: 7564.996 ymin: 530635.8 xmax: 468754.8 ymax: 1218625
## Projected CRS: OSGB36 / British National Grid
# Join spatial data with falls_admissions_75_summary
mapped_data <- combined_pop_simd2 %>%
  full_join(NHS_healthboards,by="h_bname") %>% 
  st_as_sf()
#CLAUDE : 
  library(ggiraph)
plot_map <- mapped_data %>% 
  ggplot() + 
  geom_sf_interactive(  # Changed from geom_sf
    aes(fill = quantity_per_head,
        tooltip = paste0(h_bname, 
                        "\nWinter Year: ", winter_year,
                        "\nQuantity per Head: ", round(quantity_per_head, 2))),
    colour = "white", 
    size = 0.1
  ) + 
  scale_fill_distiller(palette = "Blues", direction = 1, 
                       name = "Items per Head") +
  labs(
    title = "Antidepressant Prescriptions per Head",
    subtitle = "By Health Board and Winter Year"
  ) +
  facet_wrap(~ winter_year) +
  theme_void() +
  theme(
    strip.text = element_text(size = 12, face = "bold"),
    plot.title = element_text(face = "bold", size = 16),
    plot.subtitle = element_text(size = 10)
  )

interactive_map <- girafe(ggobj = plot_map)  # Changed from ggplotly
interactive_map

retry making boxplot

files <- list.files(here("data", "winter_population"), pattern = "csv")
winter_population <- files %>% 
  map_dfr(~read_csv(here("data", "winter_population", .))) %>% 
clean_names()

summary_combined <- combined_pop_simd %>% 
  group_by(simd2020v2_quintile,h_bname,gp_practice,winter_year) %>% 
  summarise(avg_paid_over_winters = mean(sum(paid_quantity))) %>% 
    filter(!is.na(simd2020v2_quintile))

files <- list.files(here("data", "winter_population"), pattern = "csv")
winter_population <- files %>% 
  map_dfr(~read_csv(here("data", "winter_population", .))) %>% 
clean_names()

filtered_winter_population <- winter_population %>% 
  select(date,practice_code,hb,all_ages) %>% 
  mutate(year = as.numeric(substr(date,1,4))) %>% 
  rename(healthboards = hb) %>% 
rename(winter_year = year)
 filtered_winter_population$winter_year <- gsub("2020","2019",filtered_winter_population$winter_year) 
  
  filtered_winter_population$winter_year <- gsub("2023","2022",filtered_winter_population$winter_year) 

  summary_winter_with_simd <- summary_combined %>% 
  filter(winter_year %in% c("2019","2022")) %>% 
  rename(practice_code = gp_practice) %>% 
  rename(winter_year = winter_year)
  
  summary_winter_with_simd <- summary_winter_with_simd %>%
  mutate(winter_year = as.numeric(winter_year))

filtered_winter_population <- filtered_winter_population %>%
  mutate(winter_year = as.numeric(winter_year))

joined_data <- summary_winter_with_simd %>% 
  full_join(filtered_winter_population,join_by (winter_year == winter_year, practice_code == practice_code))

final_boxplot <- joined_data %>% 
  group_by(simd2020v2_quintile,h_bname,practice_code,winter_year) %>% 
  reframe(avg_per_1000=(avg_paid_over_winters/all_ages)*1000,
            year_label=factor(winter_year,
                           level = c(2019,2022),
                           labels = c("Pre-COVID (2019)","Post-COVID (2022)")))

#chat 
hb_points <-final_boxplot %>% 
  group_by(simd2020v2_quintile,h_bname) %>%   summarise(hb_avg = mean(avg_per_1000, na.rm = TRUE))

#chat 

final_boxplot <- final_boxplot %>% 
  mutate(simd2020v2_quintile = factor(simd2020v2_quintile))
hb_points <- hb_points %>% 
  mutate(simd2020v2_quintile = factor(simd2020v2_quintile))


box2 <- final_boxplot %>%
  ggplot(aes(x = factor(simd2020v2_quintile), y = avg_per_1000, 
             fill = factor(simd2020v2_quintile))) +
  
  geom_boxplot(outlier.shape = NA) +   # hide boxplot outliers
  geom_point(data = hb_points,
             aes(x = factor(simd2020v2_quintile), y = hb_avg,
                 text = paste0("Health Board: ", h_bname, "<br>",
                               "SIMD: ", simd2020v2_quintile, "<br>",
                               "Avg per 1000: ", round(hb_avg, 1))),
             color = "red",
             size = 2,
             alpha = 0.7,
             position = position_jitter(width = 0.2)) +
  
  scale_fill_viridis(discrete = TRUE, alpha = 0.6) +
  scale_y_continuous(labels = scales::label_comma()) +
  coord_cartesian(ylim = c(5000, 20000000)) +
  
  theme_minimal() +
  theme(legend.position = "none") +
  facet_wrap(~winter_year) +
  labs(title = "Average Winter Antidepressant Prescriptions Per Practice by SIMD Quintile",
       x = "SIMD Quintile (1 = Most Deprived)",
       y = "Avg prescriptions per 1000")

# Interactive plot
ggplotly(box2, tooltip = "text")
#just boxplot

box2 <- final_boxplot %>% 
  ggplot(aes(x=simd2020v2_quintile,y=avg_per_1000, fill= factor(simd2020v2_quintile)))+
geom_boxplot() + geom_point(data = hb_points,
           aes(x = factor(simd2020v2_quintile),
               y = hb_avg),
           color = "red",
           size = 2,
           alpha = 0.7,
           position = position_jitter(width = 0.2)) +
  scale_fill_viridis(discrete = TRUE, alpha=0.6)+
  scale_y_continuous(labels = scales::label_comma()) +
  coord_cartesian(ylim = c(5000,20000000))+
  geom_jitter_interactive(color = "lightblue",size=0.4,alpha=0.9)+
  theme_minimal()+
  theme(legend.position = "none")+
  facet_wrap(~winter_year) + 
  labs(title = "Average Winter Antidepressant Prescriptions Per by SIMD Quintile",
    x = "SIMD Quintile (1 = Most Deprived)",
    y = "Avg prescriptions per 1000")

ggplotly(box2)

retry lollipop

lollipop_data <- combined_pop_simd %>% 
  filter(winter_year %in% c("2019","2022")) %>% 
  group_by(simd2020v2_quintile,h_bname,winter_year) %>% 
    summarise(total_paid = sum(paid_quantity,na.rm=TRUE),avg_SIMD = mean(simd2020v2_quintile, na.rm=TRUE)) %>% 
mutate(period=ifelse(winter_year==2019, "pre", "post")) 

#cht

lollipop1_summary <- lollipop_data %>%
  
  # Step 1: Aggregate prescribing for each HB × period
  group_by(h_bname, period) %>%
  summarise(
    total_prescribing = sum(total_paid, na.rm = TRUE),
    avg_simd = mean(simd2020v2_quintile, na.rm = TRUE),
    .groups = "drop"
  ) %>%
  
  # Step 2: Remove HBs with missing names
  filter(!is.na(h_bname)) %>%
  
  # Step 3: Pivot pre/post
  tidyr::pivot_wider(
    names_from = period,
    values_from = total_prescribing
  ) %>%
  
  # Step 4: Calculate percentage change
  mutate(
    pct_change = ((post - pre) / pre) * 100
  ) %>%
  
  # Step 5: Rank health boards by average SIMD
  arrange(avg_simd) %>%
  mutate(h_bname = factor(h_bname, levels = unique(h_bname)))


pct_lollipop <-lollipop1_summary %>% 
  ggplot () +
  geom_segment(aes(x=h_bname, xend=h_bname, y=0, yend = pct_change), color = "skyblue")+
  geom_point(aes(x = h_bname, y = pct_change),color = "blue", size=3, alpha = 0.6) +
  theme_light() +
  coord_flip() +  
  theme(panel.grid.major.y = element_blank(),
    panel.border = element_blank(),
    axis.ticks.y = element_blank()) +
  labs( title = "Percentage Change in Prescribing per Health Board",
    x = "Health Board",
    y = "Percentage Change (%)")
   ggplotly(pct_lollipop)

questions : not sure the best way to displa my original data ? github - how to get rid of the signs 1- overall national trend (use original graph) 2- i want to show the variation between different regions using healthboards 3- link it to deprevation and look at prescriptions per person 4- can i do a map that shows pre covid and post covid side by side would that count as one 5- voilin plot across differet SIMDs to compare smaller unit of data - gp practice (postcode that links to SIMD) PATCHWORK - MAPS TOGETHETE reference line to show the split between precovid, covid and postcovid

every dot is a gp practice - gp practice - dataset - adressess (assessment prep) voilin plot if messy add transperency open data use quintiles for voilin plot do a code that says if not installed install and load packages

percentage increase

overall trend map antidepressant prescribing per head by healthboard facet by winter year? boxplot by SIMD dumbell plot / lollipop graph - percentage change write where the links are from